Abstract
Abstract:
Several national total joint arthroplasty registries exist outside of the United States (U.S.) and have been used to compare rates and outcomes of total knee arthroplasty. Within the U.S., regional arthroplasty registries provide an opportunity to compare U.S. practices and outcomes with those of other countries. The purpose of this study was to compare the demographics, choice of implants, techniques, and outcomes of total knee arthroplasties in Norway to those from a large, U.S. integrated health-care system and to determine the feasibility of using aggregate-level data for international registry comparisons. The study sample consisted of 25,004 primary total knee arthroplasties performed in Norway and 56,208 from the Kaiser Permanente health-care system. Summary-level data were used to compare the two cohorts. At the time of the seven-year follow-up, the cumulative survival of the total knee prosthesis was 94.8% for the arthroplasties performed in Norway and 96.3% for those performed at Kaiser Permanente. The primary reasons for revision arthroplasty included infection, instability, pain, and aseptic loosening. Patient characteristics, selection of implants, surgical techniques, and outcomes differed between the cohorts. Harmonization of data elements and definitions is necessary for future international research.
Several national total joint arthroplasty registries have been influential in reducing the revision rates associated with total joint arthroplasty by identifying implant failures and providing feedback on clinical practices1-7. Comparisons between these national registries have provided important information on variation in incidence rates, surgical techniques, implant selection, and outcomes of total joint arthroplasty procedures8,9.
Within the U.S., claims and administrative databases have been used to compare arthroplasty rates, demographics, and revision rates with those of other countries10. While these studies provide important information on total joint arthroplasty revision rates through the use of large, representative samples, they are limited by the lack of laterality and specific clinical and implant details that are necessary for assessment of implant survival after total joint arthroplasty.
Total joint arthroplasty registries provide an alternative to the use of claims data for the assessment of total joint arthroplasty outcomes in the U.S. Several regional and institutional registries exist within the U.S.11-13 and collect the data (e.g., patient demographics, surgical technique, type of implant, and reasons for revision) necessary to assess total joint outcomes. Although these registries have been used to assess total joint replacement outcomes within specific institutions and/or regions, direct comparisons with international total joint arthroplasty registries have not been published, to our knowledge.
These regional and institutional total joint arthroplasty registries provide a unique opportunity to compare U.S. total joint arthroplasty demographics, practices, and revision arthroplasty rates with those of other countries. Therefore, the purpose of this study was to compare patient characteristics, implant characteristics, surgical techniques, revision rates, and reasons for revision arthroplasty in a cohort of patients with primary total knee arthroplasty from the Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry to assess the feasibility of aggregating international registry data while identifying the strengths and limitations of such an approach.
Institutional review board approval was obtained for this study, allowing for the sharing of summary data while protecting individual patient health information at both institutions.
Data Sources
The Norwegian Arthroplasty Register, a national registry that was started in 1987 as a hip registry, contains data regarding more than 114,400 hip replacements. The Norwegian knee registry was established in 1994 and, by 2009, had registered data on 40,000 knee arthroplasties. The registry tracks total knee arthroplasty procedures in a population of approximately 4.6 million2,14, with high completeness of reporting15. The Kaiser Permanente Total Joint Replacement Registry was developed in 2001 and registered over 63,000 knee arthroplasty cases by the end of 2009. Kaiser Permanente provides health-care coverage for 8.6 million members who are part of a large, integrated health-care system in seven geographical areas of the U.S.16, and it has reported good participation and completion rates with regard to the registry13. The Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry were used to identify primary total knee arthroplasty cases that were performed between January 1, 2001, and December 31, 2009. Patient demographics, surgical techniques, type of implants, cumulative survival rates, and reasons for revision arthroplasty were summarized from each registry by means of tables with summary level statistics (percent, means, medians, and standard deviation) and survival function by strata. Revision for all reasons, aseptic revision only (excluding revision due to infection), and revision due to infection were reported.
Data Collection
All data are reported to the Norwegian Arthroplasty Register by the orthopaedic surgeon, who completes the same standard paper form for each primary or revision arthroplasty (but with separate forms used for hip replacements and for replacements of joints other than the hips)15. Similarly, operative data are documented by the surgeon at the point of care in the Kaiser Permanente Total Joint Replacement Registry. In addition to operative forms, preoperative and postoperative ambulatory encounters are also captured with use of electronic forms in the Kaiser Permanente Total Joint Replacement Registry. These forms are then supplemented with additional data elements from the electronic medical record of the patients13. The Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry collect similar data elements, including patient demographics, implant names and attributes, surgical techniques, revisions, and reasons for revisions. Some differences in documentation exist. For example, American Society of Anesthesiologists (ASA) classification in the Norwegian Arthroplasty Register is assigned by surgeons, whereas in the Kaiser Permanente Total Joint Replacement Registry this classification is determined by anesthesiologists.
Statistical Analyses
Descriptive statistics such as means, standard deviations, and proportions were used to describe the study sample. Chi square, Fisher exact test, and independent t tests were applied to compare demographics, surgical techniques, types of implants, revision rates, and reasons for revision arthroplasty. Kaplan-Meier survival curves with revision as the end point were used to compare implant survival from the different registries. Revisions in the Norwegian Arthroplasty Register were defined as reoperations in which implant parts were added, exchanged, or removed, and the date of the first revision (when the procedure involves a two-stage revision) was considered to be the revision date. In the Kaiser Permanente Total Joint Replacement Registry, revisions were defined as a reoperation in which any implant was exchanged and/or added. If a two-stage revision was performed (i.e., the prosthesis was removed during the first operation and a new prosthesis was implanted during the second), the date of the second procedure was considered to be the revision date. If no second operation was performed (for instance, if the patient died after the first procedure or if the second procedure could not be performed for any other reason), then the date of the first-stage revision procedure was considered to be the revision date.
Patients who died, emigrated, left the hospital health plan, or reached the end of the study period without a reported outcome were censored in the survival analyses. SAS software (version 9.1.3 for Windows; SAS Institute, Cary, North Carolina) and SPSS software (SPSS for Windows, release 18.0; SPSS, Chicago, Illinois) were used to analyze the data, with p < 0.05 used as the statistical threshold.
The study consisted of 25,004 primary total knee arthroplasties performed in Norway and 56,208 primary total knee arthroplasties performed at Kaiser Permanente. During the study period, 9.1% of the patients died and 0.2% of the patients were lost to follow-up in the Norwegian cohort. Within the Kaiser Permanente Total Joint Replacement Registry cohort, 3.4% of the patients died and 7.5% of the patients left the Kaiser Permanente health plan during the study period.
Patient Characteristics and Diagnosis
Comparisons of patient demographics indicated that the Norwegian total knee arthroplasty patients were older (p < 0.001), had a lower proportion of osteoarthritis as a diagnosis (p < 0.001), and had lower ASA scores (p < 0.001) as compared with the Kaiser Permanente patients (Table I). Kaiser Permanente also had a significantly higher proportion of male patients who received total knee arthroplasty than Norway (p < 0.001).
Fixation and Implant Type
Implant fixation differed between Norway and Kaiser Permanente, with a significantly higher rate of uncemented total knee arthroplasty fixation in the Kaiser Permanente Total Joint Replacement Registry (p < 0.001) (Table II). The types of femoral components also differed between Norway and Kaiser Permanente. In Norway, Profix (Smith & Nephew, London, United Kingdom) (37%), Low Contact Stress (LCS, DePuy Orthopaedics, Warsaw, Indiana) (12.8%), LCS Complete (DePuy Orthopaedics) (20.6%), and AGC (Biomet, Bridgend, South Wales, United Kingdom) (11.6%) accounted for the majority of total knee arthroplasty implants, while Press-Fit Condylar (PFC, DePuy Orthopaedics) (48.0%) and NexGen (Zimmer, Warsaw, Indiana) (39.0%) were the primary implants used at Kaiser Permanente.
The use of mobile-bearing total knee prostheses also differed between Norway and Kaiser Permanente. A low-contact-stress design was used in 33% of the total knee arthroplasties in Norway, and mobile-bearing knees were used in 35% of the knees. In the Kaiser Permanente Total Joint Replacement Registry, mobile-bearing knees represented 9.6% of total knee arthroplasty implant usage, with the majority of implants being the PFC-Rotating Platform implant.
The use of cruciate-retaining or substituting fixed-bearing total knee arthroplasty knee implants differed between Norway and Kaiser Permanente. The majority of fixed-bearing total knee arthroplasty implants used in Norway were cruciate-retaining implants (60.9% versus 3.5%), whereas a posterior-stabilized design was used in the majority of Kaiser Permanente patients (56.4% versus 29.6%), p < 0.001.
In comparing patellar resurfacing, there was also a significant difference in practice between Norway and Kaiser Permanente. While 94.7% of total knee arthroplasties were performed without resurfacing of the patella in Norway, 98.3% of total knee arthroplasties were performed with patellar resurfacing at Kaiser Permanente.
Operative Time
Mean operative time and standard deviation (from the time of the first incision to the time of completion of skin closure) for total knee arthroplasties was slightly higher for the Kaiser Permanente patients compared with the Norwegian patients (96.4 min ± 33.3 versus 95.5 min ± 31.8, p < 0.001). While this difference is significant, clinical significance is questionable.
Revision Rates and Reasons for Revision
The overall cumulative survival of total knee implants at seven years was 94.8% (95% confidence interval [CI], 94.4% to 95.2%) for Norway and 96.3% (95% CI, 96.0% to 96.6%) for Kaiser Permanente. For aseptic loosening, cumulative survival was lower for Norway than for Kaiser Permanente. While cumulative total knee implant survival was higher for Kaiser Permanente when aseptic loosening was used as the end point, Norway had a higher cumulative survival when infection was used as the end point. Comparisons of specific implants indicated that the LCS was associated with a higher revision rate for both Kaiser Permanente and Norway (Figs. 1 through 8).
Reasons for revision during the study differed between Kaiser Permanente and Norway (Table III). While total knee arthroplasty revisions due to infection were more common in the Kaiser Permanente patients than in the Norwegian patients (p < 0.001), pain was the most common reason for total knee arthroplasty revision in Norway (p < 0.001). Aseptic loosening as a reason for revision arthroplasty was more common in the Norwegian patients than it was in the Kaiser Permanente patients (p < 0.001). Arthrofibrosis as a reason for revision was higher in the Kaiser Permanente patients than it was in the Norwegian patients (10.8% versus 3.9%).
Comparisons between the Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry suggest that these primary total knee arthroplasty cohorts differ with regard to patient demographics, implants, surgical techniques, and survival of the total knee arthroplasty implant. While mean age was similar, distributions were slightly different between Kaiser Permanente and Norway, with patients in the Norwegian Arthroplasty Register being older. Both the Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry had a predominance of female patients, with the Kaiser Permanente Total Joint Replacement Registry having a slightly higher percentage of male patients compared with that in the Norwegian Arthroplasty Register.
The apparent mild disparity between patients in the Norwegian Arthroplasty Register and patients in the Kaiser Permanente Total Joint Replacement Registry with regard to the prevalence of osteoarthritis as the underlying diagnosis may represent a lag in the demand for total knee arthroplasty for this condition in Norway. Due to earlier adoption of total knee arthroplasty for the treatment of knee osteoarthritis in the U.S., the Kaiser Permanente Total Joint Replacement Registry probably represents a more mature state in which the demand for total knee arthroplasty for the treatment of end-stage arthritis has been met. Such an artifact could skew the findings toward a greater proportion of total knee arthroplasty performed for rheumatoid arthritis and inflammatory arthritis in the Norwegian register8.
Differences in ASA scores between the Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry may suggest higher comorbidities in the Kaiser Permanente Total Joint Replacement Registry cohort. With an older patient population, ASA scores would be expected to be elevated in Norway; however, obesity and other comorbidities may be higher in the Kaiser Permanente cohort and may explain the higher ASA scores. The difference in ASA scores could also be accounted for by interreliability issues. In Norway, the surgeon assigns the ASA designation. In the United States, patients are typically classified by anesthesiologists, who may overestimate ASA scores. Studies that have assessed both intrarater and interrater reliability of ASA scores indicate low reliability. Other comorbidity indices, such as the Charlson score17, may provide more reliable measures of patient comorbidities for future studies.
Several differences in surgical techniques were also observed between the two cohorts. Comparisons of the Norwegian Arthroplasty Register and Kaiser Permanente Total Joint Replacement Registry indicated differences in patellar resurfacing and fixation methods. More than 94% of total knee arthroplasties were performed without resurfacing of the patella in the Norwegian cohort, whereas more than 98% of total knee arthroplasties were performed with patellar resurfacing in the Kaiser Permanente cohort. Many of the revision procedures in the Norwegian cohort were performed for anterior knee pain in patients who had undergone total knee arthroplasty without initial patellar resurfacing. Although these procedures were technically reoperations, the femoral and tibial implants were not revised in those limbs. While Norway and Sweden have similar practices with regard to nonpatellar resurfacing, Denmark, similar to the U.S., has a practice for patellar resurfacing8. In Norway, the functional status and revision rate associated with patellar resurfacing and nonpatellar resurfacing total knee arthroplasty was investigated and no difference was encountered, which may serve as an explanation for the low use of patellar resurfacing in that country18,19. A slightly higher revision rate in nonresurfaced knees might be explained by the option of adding a patellar component in nonresurfaced knees19.
In addition to differences in surgical techniques, we also observed differences in implant selection between the two cohorts. There was significant variation in the types of implant used in each country, with few implants overlapping between the cohorts. Implant designs also differed, with mobile-bearing total knee prostheses implanted at a higher frequency in the Norway cohort than in the Kaiser Permanente cohort. An LCS design was used in more than 33.4% of total knee arthroplasty procedures in the Norwegian cohort. In the Kaiser Permanente cohort, mobile-bearing knees were used in almost 10% of the total knee arthroplasty procedures, with the majority of the prostheses being the PFC posterior-stabilized (PS) implant. Initial results suggest that the LCS was associated with a higher revision rate in both registries. A more in-depth assessment of these mobile-bearing implants is planned as part of the International Consortium of Orthopaedic Registries (ICOR) initiative.
The use of cruciate-retaining or cruciate-substituting fixed-bearing total knee arthroplasty knee designs also differed between the Norwegian Arthroplasty Register and the Kaiser Permanente Total Joint Replacement Registry. The majority of fixed-bearing total knee arthroplasty procedures in Norway involved cruciate-retaining implants, whereas, in the Kaiser Permanente system, a posterior-stabilized design was used in the majority of cases. There has long been a debate on the merits of each design, and this debate cannot be addressed within this observational study due to the low number of cases in which a posterior-stabilized design was used in Norway.
Cumulative Survival of Total Knee Implants and Reasons for Revision
Differences in patient characteristics were controlled for by stratification on key variables to assess the survival of total knee implants. Similar to the findings of Kurtz et al.10, the overall cumulative survival of total knee implants was higher in the Kaiser Permanente cohort than it was in the Norwegian cohort. However, in assessing infection as the end point, a higher cumulative survival rate was observed for the Norwegian total knee implants. Identification of knee infections after total knee arthroplasty may be underreported in the Norwegian registry, however, since only the reoperations in which an implant is removed or exchanged are identified. The AGC implant was used in approximately 10% of total knee arthroplasty procedures in Norway. Knee infections after AGC total knee implantation might not be captured in the data as deep infection, even though an operative irrigation and debridement is performed. The reason for this is that AGC monoblock tibial components do not have an exchangeable tibial liner, and thus this surgical procedure would not be categorized as a revision procedure. Within the Kaiser Permanente organization, a comprehensive infection surveillance protocol with chart review captures occurrence of deep infection whether a revision was performed or not. Another potential explanation is the difference in the use of antibiotic-loaded bone cement. While antibiotic-loaded bone cement has not been shown to reduce infection rates in U.S. patients after primary total knee arthroplasty20,21, the use of antibiotic-loaded bone cement was substantially different between the two cohorts in our study. Antibiotic-loaded bone cement was used in 100% of the Norwegian primary total knee arthroplasties, but only in 12% of primary total knee arthroplasties in the Kaiser Permanente cohort. The main reason for the high use of antibiotic-loaded bone cement in Norway is because a lower rate of revisions due to infections was found after total hip replacement in a Norwegian cohort when antibiotic-loaded bone cement was used22. The infection rate was also shown to be lower after total knee arthroplasty in a Finnish cohort of patients in whom antibiotic-loaded cement was used23,24. In both registries (Kaiser Permanente and Norway), intravenous prophylactic antibiotics are routinely administered to all patients undergoing total knee arthroplasty (Table IV). Additional analyses are necessary to identify the underlying source of these differences in infection.
Revision arthroplasty surgery is considered to be an important outcome in the surveillance of primary total knee arthroplasty procedures by both registries. Interpretation of revision rates, however, must be interpreted with care because of differences in “community standards” in Norway and the United States. These important differences include the proportion of total knee arthroplasty procedures that are performed without resurfacing the patella and the use of unicompartmental knee replacements. Differences in data collection methods and definitions are also critical to the interpretation of study findings.
Strengths and Limitations of the Study
This study has several strengths and limitations. The strengths of the study include the large sample sizes from established registries with prospectively collected data and contemporary implants. The limitations include the observational research design, the short to intermediate period of follow-up, and the lack of patient-reported outcomes. Other limitations include the lack of standardization of reporting the diagnosis, the different choices of diagnosis, and the cause of revision on the patient forms. The different number of patients lost to follow-up in the two cohorts might also influence the reported revision rate.
Conclusions
Total joint arthroplasty registries provide an important role in post-market surveillance of total knee arthroplasty implants. The comparisons made between the Norwegian knee arthroplasty register and the Kaiser Permanente registry highlight important similarities and differences between the outcomes of total knee arthroplasty, and the surgical practices, in each country. This study also emphasizes the need to address regional and national differences in demographics, surgical techniques, implants, and definitions in order to compare results across existing registries.
Understanding the differences in surgical practices was recognized as an important factor in the interpretation of the data. We have identified areas of interest for focused analyses and hope to expand this collaboration. Development of collaborations via a global network of international registries such as ICOR will require development of structured common data elements, convergence of definitions for outcomes, recognition of differences in patient populations and surgical practices, and facilitation of the removal of regulatory and legal obstacles to enhance future orthopaedic research efforts.
Graves
SE;
Davidson
D;
Ingerson
L;
Ryan
P;
Griffith
EC;
McDermott
BF;
McElroy
HJ;
Pratt
NL. The Australian Orthopaedic Association National Joint Replacement Registry. Med J Aust.
2004;180(
5 Suppl):S31-4.[PubMed]
Havelin
LI;
Engesaeter
LB;
Espehaug
B;
Furnes
O;
Lie
SA;
Vollset
SE. The Norwegian Arthroplasty Register: 11 years and 73,000 arthroplasties. Acta Orthop Scand.
2000;71:337-53.[PubMed][CrossRef]
Herberts
P;
Malchau
H. Long-term registration has improved the quality of hip replacement: a review of the Swedish THR Register comparing 160,000 cases. Acta Orthop Scand.
2000;71:111-21.[PubMed][CrossRef]
Kolling
C;
Simmen
BR;
Labek
G;
Goldhahn
J. Key factors for a successful National Arthroplasty Register. J Bone Joint Surg Br.
2007;89:1567-73.[PubMed][CrossRef]
Lucht
U. The Danish Hip Arthroplasty Register. Acta Orthop Scand.
2000;71:433-9.[PubMed][CrossRef]
Malchau
H;
Herberts
P;
Ahnfelt
L. Prognosis of total hip replacement in Sweden. Follow-up of 92,675 operations performed 1978-1990. Acta Orthop Scand.
1993;64:497-506.[PubMed][CrossRef]
Puolakka
TJ;
Pajamäki
KJ;
Halonen
PJ;
Pulkkinen
PO;
Paavolainen
P;
Nevalainen
JK. The Finnish Arthroplasty Register: report of the hip register. Acta Orthop Scand.
2001;72:433-41.[PubMed][CrossRef]
Robertsson
O;
Bizjajeva
S;
Fenstad
AM;
Furnes
O;
Lidgren
L;
Mehnert
F;
Odgaard
A;
Pedersen
AB;
Havelin
LI. Knee arthroplasty in Denmark, Norway and Sweden. Acta Orthop.
2010;81:82-9.[PubMed][CrossRef]
W-Dahl
A;
Robertsson
O;
Lidgren
L;
Miller
L;
Davidson
D;
Graves
S. Unicompartmental knee arthroplasty in patients aged less than 65. Acta Orthop.
2010;81:90-4.[PubMed][CrossRef]
Kurtz
SM;
Ong
KL;
Schmier
J;
Mowat
F;
Saleh
K;
Dybvik
E;
Kärrholm
J;
Garellick
G;
Havelin
LI;
Furnes
O;
Malchau
H;
Lau
E. Future clinical and economic impact of revision total hip and knee arthroplasty. J Bone Joint Surg Am.
2007;89
Suppl 3:144-51.[PubMed][CrossRef]
Berry
DJ;
Kessler
M;
Morrey
BF. Maintaining a hip registry for 25 years. Mayo Clinic experience. Clin Orthop Relat Res.
1997;344:61-8.[PubMed][CrossRef]
Gioe
TJ;
Killeen
KK;
Mehle
S;
Grimm
K. Implementation and application of a community total joint registry: a twelve-year history. J Bone Joint Surg Am.
2006;88:1399-404.[PubMed][CrossRef]
Paxton
EW;
Inacio
MC;
Khatod
M;
Yue
EJ;
Namba
RS. Kaiser Permanente National Total Joint Replacement Registry: aligning operations with information technology. Clin Orthop Relat Res.
2010;468:2646-63.[PubMed][CrossRef]
Furnes
O;
Espehaug
B;
Lie
SA;
Vollset
SE;
Engesaeter
LB;
Havelin
LI. Early failures among 7,174 primary total knee replacements: a follow-up study from the Norwegian Arthroplasty Register 1994-2000. Acta Orthop Scand.
2002;73:117-29.[PubMed][CrossRef]
Espehaug
B;
Furnes
O;
Havelin
LI;
Engesaeter
LB;
Vollset
SE;
Kindseth
O. Registration completeness in the Norwegian Arthroplasty Register. Acta Orthop.
2006;77:49-56.[PubMed][CrossRef]
Paxton
EW;
Namba
RS;
Maletis
GB;
Khatod
M;
Yue
EJ;
Davies
M;
Low
RB
Jr;
Wyatt
RW;
Inacio
MC;
Funahashi
TT. A prospective study of 80,000 total joint and 5000 anterior cruciate ligament reconstruction procedures in a community-based registry in the United States. J Bone Joint Surg Am.
2010;92
Suppl 2:117-32.[PubMed][CrossRef]
Charlson
ME;
Pompei
P;
Ales
KL;
MacKenzie
CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis.
1987;40:373-83.[PubMed][CrossRef]
Lygre
SH;
Espehaug
B;
Havelin
LI;
Vollset
SE;
Furnes
O. Does patella resurfacing really matter? Pain and function in 972 patients after primary total knee arthroplasty. Acta Orthop.
2010;81:99-107.[PubMed][CrossRef]
Lygre
SH;
Espehaug
B;
Havelin
LI;
Vollset
SE;
Furnes
O. Failure of total knee arthroplasty with or without patella resurfacing. Acta Orthop.
2011;82:282-92.[PubMed][CrossRef]
Namba
RS;
Chen
Y;
Paxton
EW;
Slipchenko
T;
Fithian
DC. Outcomes of routine use of antibiotic-loaded cement in primary total knee arthroplasty. J Arthroplasty.
2009;24(
6 Suppl):44-7.[PubMed][CrossRef]
Gandhi
R;
Razak
F;
Pathy
R;
Davey
JR;
Syed
K;
Mahomed
NN. Antibiotic bone cement and the incidence of deep infection after total knee arthroplasty. J Arthroplasty.
2009;24:1015-8.[PubMed][CrossRef]
Engesaeter
LB;
Lie
SA;
Espehaug
B;
Furnes
O;
Vollset
SE;
Havelin
LI. Antibiotic prophylaxis in total hip arthroplasty: effects of antibiotic prophylaxis systemically and in bone cement on the revision rate of 22,170 primary hip replacements followed 0-14 years in the Norwegian Arthroplasty Register. Acta Orthop Scand.
2003;74:644-51.[PubMed][CrossRef]
Jämsen
E;
Furnes
O;
Engesaeter
LB;
Konttinen
YT;
Odgaard
A;
Stefánsdóttir
A;
Lidgren
L. Prevention of deep infection in joint replacement surgery. Acta Orthop.
2010;81:660-6.[PubMed][CrossRef]
Jämsen
E;
Huhtala
H;
Puolakka
T;
Moilanen
T. Risk factors for infection after knee arthroplasty. A register-based analysis of 43,149 cases. J Bone Joint Surg Am.
2009;91:38-47.[CrossRef]